Eur J Pediatr DOI 10.1007/s00431-011-1461-2

ORIGINAL PAPER

Lung ultrasound in bronchiolitis: comparison with chest X-ray Vito Antonio Caiulo & Luna Gargani & Silvana Caiulo & Andrea Fisicaro & Fulvio Moramarco & Giuseppe Latini & Eugenio Picano

Received: 14 February 2011 / Accepted: 22 March 2011 # Springer-Verlag 2011

Abstract The diagnosis of bronchiolitis is based mainly on the patient’s medical history and physical examination. However, in severe cases, a further evaluation including chest X-ray (CXR) may be necessary. At present, lung ultrasound (LUS) is not included in the diagnostic work-up of bronchiolitis. This study aimed to compare the diagnostic accuracy of LUS and CXR in children with bronchiolitis, and to evaluate the correlation between clinical and ultrasound findings. Only patients with a diagnosis of bronchiolitis, who had undergone a CXR, were enrolled in the study. Fifty-two infants underwent LUS and CXR. LUS was also performed in 52 infants without clinical signs of bronchiolitis. LUS was positive for the diagnosis of bronchiolitis in 47/52 patients, whereas CXR was positive in 38/52. All patients with normal LUS examination had a

normal CXR, whereas nine patients with normal CXR had abnormal LUS. In these patients, the clinical course was consistent with bronchiolitis. We found that LUS is a simple and reliable tool for the diagnosis and follow-up of bronchiolitis. It is more reliable than CXR, can be easily repeated at the patient’s bedside, and carries no risk of irradiation. In some patients with bronchiolitis, LUS is able to identify lung abnormalities not revealed by CXR. Furthermore, there is a good correlation between clinical and ultrasound findings. Given the short time needed to get a US report, this technique could become the routine imaging modality for patients with bronchiolitis.

V. A. Caiulo (*) : F. Moramarco Department of Pediatrics, Perrino Hospital, Piazza Angeli, 3-72100 Brindisi, Italy e-mail: [email protected]

Abbreviations LUS Lung ultrasound B Bronchiolitis CXR Chest X-ray

L. Gargani : E. Picano Institute of Clinical Physiology, National Council of Research, Division of Pisa, Pisa, Italy S. Caiulo : A. Fisicaro School of Medicine, San Raffaele University, Milan, Italy G. Latini Department of Neonatology, Perrino Hospital, Brindisi, Italy G. Latini Institute of Clinical Physiology, National Council of Research, Division of Lecce, Lecce, Italy

Keywords Lung ultrasound . Bronchiolitis . B-lines . Chest X-ray

Introduction Bronchiolitis is an acute respiratory illness that particularly affects children younger than 24 months of age, with severe disease more common among infants aged 1–3 months. The first signs of infection are coryza and sometimes low-grade fever that progress over a few days to cough, tachypnea, hyperinflation, chest retraction, and widespread crackles, wheezes, or both. Bronchiolitis is a clinical diagnosis. Routine laboratory tests offer little helpful information in most cases. Abnormalities on chest X-ray (CXR) range from 20% to 96% [4].

Eur J Pediatr

The use of ultrasound for studying the lung is relatively recent. Lung ultrasound (LUS) has proved to be a useful tool for evaluating many different cardiopulmonary conditions [14, 15, 17–19, 22, 25]. LUS is especially valuable since it is a very easy application of echography; it is rapid, portable, repeatable, and non-ionizing. This last issue is especially important in infants, who carry a higher risk of cancer from exposure to radiation, than people of other ages [20]. Therefore, alternative diagnostic methods that do not involve the use of ionizing radiation should be considered when evaluating young individuals, in order to minimize cancer risk [6]. In the normal lung, which consists mostly of air, ultrasound waves are reflected, without being translated into an image. The pleura is the only visible structure since the high-acoustic impedance of the air below prevents visualization of the lung parenchyma. Any process that leads to the reduction of the air content, such as the presence of extravascular lung water or a lung consolidation process, allows the reflection of the ultrasound beam and the evaluation of the diseased lung. In the absence of any lung parenchyma disease, reverberation artifacts, repetitive and parallel to the pleura, called Alines, are generated. In the presence of pathological processes that lead to thickening of peripheral interlobular septa, A-lines are replaced by other comet-tail artifacts, perpendicular to the pleural line, called B-lines (also known as ultrasound lung comets) [16]. Their physical genesis is linked to the thickening of the interlobular interstitium of subpleural secondary lobules, resulting in the reverberation of the ultrasound beam [21]. When an infectious process replaces the alveolar air content with exudate, the lung parenchyma becomes accessible to the ultrasound beam and appears similar to that of a parenchymal organ. If it comes into contact with the pleural line, is it possible to visualize the lesion, which has a random conformation with edges that are irregular, blurred, and sometimes indistinct [10]. The physical limits of ultrasound do not allow visualization of the normal, aerated lung, but may provide much useful clinical information in the diseased lung. This has led to a delay in the development of a branch of ultrasound that has only recently begun to display its potential. At present, LUS is not included in the diagnostic workup of infant respiratory diseases. Only a few studies have addressed this topic, although with striking results. In particular, LUS seems to be a valuable tool in the diagnosis of transient tachypnea of the newborn [9], respiratory distress syndrome [11], and pneumonia [10]. The aim of the study was to compare the diagnostic accuracy of ultrasound and CXR in children with bronchiolitis, and to evaluate the correlation between clinical and ultrasound findings

Methods Study population From December 2009 to December 2010, 52 patients with bronchiolitis (28 males), who had undergone a CXR for clinical reasons were enrolled in the study. The diagnosis of bronchiolitis was made according to the American Academy of Pediatrics, on the basis of history and clinical examination [24]. Ages ranged from 1 to 16 months (median 2.1 months, interquartile range 1.5–5.4). LUS was also performed in 52 infants (28 males) without clinical signs of bronchiolitis (control group), admitted for suspected gastroesophageal reflux disease; ages ranged from 1 to 22 months (median 3 months, interquartile range 1.7–5.9). In all patients, a LUS examination was performed on the day of the CXR examination. Informed consent was obtained from all parents, and the investigation followed the guidelines of the local ethics committee. The sonographer was informed about the clinical indication but he was blinded to radiographic findings. LUS was then repeated between days 3 and 5 (n=50), days 6 and 9 (n=38), and days 10 and 12 (n=15). LUS follow-up was stopped when the sonographic appearance of the lung appeared normal. Grading of distress severity The severity of the disease was assessed by Downes' Score, as previously described (Table 1) [26]. & & &

Mild bronchiolitis: score 0–3 Moderate bronchiolitis: score 4–6 Severe bronchiolitis: score 7–10

Chest X-ray examination Posterior–anterior CXR was acquired with patients in supine position. In accordance with the British Thoracic Society guidelines, lateral radiographs were not obtained [7]. Lung ultrasound examination Transthoracic LUS examinations were performed with commercially available ultrasound machines (Toshiba Nemio) equipped with a high-resolution linear probe with frequencies ranging from 6 to 12 MHz. LUS examination consisted of both longitudinal and transversal sections. On the anterior chest, transversal sections were obtained by positioning the probe transversally to the chest, from the second to the fifth intercostal space, whereas longitudinal sections were obtained by positioning the probe longitudinally to the chest, along the

Eur J Pediatr Table 1 Downes' score (modified)

Score

0

1

2

PO2 Inspiratory breath sounds Accessory muscles used Expiratory wheezing Respiratory rate

70–100 in air Normal None None ≤40/min

≤70 in air Unequal Moderate Moderate 40–70/min

≤70 in 40% O2 Decreased to absent Maximal Marked >70/min

parasternal, mid-clavicular, anterior axillary, and mid-axillary lines. On the posterior chest, transversal sections were obtained positioning the probe on the intercostal spaces below the scapular spine, whereas longitudinal sections were obtained along the paravertebral, scapular, and posterioraxillary lines. This methodology is similar to the one previously described by Copetti et al. [9–12]. The selected setting for the ultrasound probe was the same used for soft tissue analysis, with a maximum depth of 8 cm. This setting allows scanning about the entire lung area. The intra- and inter-observer variability for LUS had been previously described as 5.1% and 7.4%, respectively [21]. LUS findings were classified according the following patterns: 1. Presence of subpleural lung consolidation, defined as subpleural areas of tissue-like or poor-echo structure, with blurred margins. 2. Presence of compact B-lines, defined as areas of white lung. 3. Pleural lines abnormalities, defined as irregular appearance of the pleural line. 4. Presence of focal multiple B-lines. 5. Normal pattern, defined as normal lung sliding with or without A-lines. Statistical analysis Continuously distributed variables were expressed as median and interquartile ranges. Categorical variables were presented as counts and percentages. The demographic variables were compared by two-tailed Student's t tests. Sonographic findings in the two groups were compared by a chi-squared test. Statistical analysis was performed using the SPSS/PC (version 13, SPSS, Chicago, IL, USA) software package. For all the statistical analyses, significance was accepted at p